Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    XML headers

    np_max-act · gemini-2.0-flash

    The neuron selectively responds to numeric literals—especially floating‐point numbers (e.g. “0.73486328125”, “2.763671875”)—appearing in the XML/HTML comments.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ické
    -0.07
     LANGUAGE
    -0.06
    ět
    -0.06
     blir
    -0.06
    увався
    -0.06
     Midwest
    -0.06
     fwd
    -0.06
     таких
    -0.06
    .Pr
    -0.06
     Op
    -0.06
    POSITIVE LOGITS
     crystall
    0.07
     Pointer
    0.07
     campaigning
    0.07
     typedef
    0.06
    ypi
    0.06
    (py
    0.06
     ration
    0.06
     disk
    0.06
     sacrificing
    0.06
     zahrani
    0.06
    Act Density 0.002%

    No Known Activations